mlsanityData to 14 Oct 2026

ClawVM: Harness-Managed Virtual Memory for Stateful Tool-Using LLM Agents

arXiv 2604.10352 ↗First posted 11 Apr 2026cs.AI cs.OS cs.SEcited by 4

On the map: LLM agents and tool use → Agent memory systems

Abstract (arXiv)

Stateful tool-using LLM agents treat the context window as working memory, yet today's agent harnesses manage residency and durability as best-effort, causing recurring failures: lost state after compaction, bypassed flushes on reset, and destructive writeback. We present textscClawVM, a virtual memory layer that manages state as typed pages with minimum-fidelity invariants, multi-resolution representations under a token budget, and validated writeback at every lifecycle boundary. Because the harness already assembles prompts, mediates tools, and observes lifecycle events, it is the natural enforcement point; placing the contract there makes residency and durability deterministic and auditable. Across synthetic workloads, 12 real-session traces, and adversarial stress tests, textscClawVM eliminates all policy-controllable faults whenever the minimum-fidelity set fits within the token budget, confirmed by an offline oracle, and adds median <50 microseconds of policy-engine overhead per turn.

Key facts. Cited by 4 papers in the map (data through 14 Oct 2026). Placed in Agent memory systems, one of 16 directions in LLM agents and tool use. Builds on 17 of its 44 references that are in the map. First posted on arXiv 11 Apr 2026.

Builds on

17 of 44 references.

Closest ancestors · 7

Same direction, newest first

  1. 2026Mem2ActBench: A Benchmark for Evaluating Long-Term Memory Utilization in Task-Oriented Autonomous Agents13 Jan
  2. 2025Memory as Action: Autonomous Context Curation for Long-Horizon Agentic Tasks14 Oct
  3. 2025Text2Mem: A Unified Memory Operation Language for Memory Operating System14 Sep
  4. 2025Evaluating Memory in LLM Agents via Incremental Multi-Turn Interactions7 Jul
  5. 2025Memory OS of AI Agent30 May
  6. 2025A-MEM: Agentic Memory for LLM Agents17 Feb
  7. 2024Evaluating Very Long-Term Conversational Memory of LLM Agents27 Feb

Other references · 10

  1. 2023Generative Agents: Interactive Simulacra of Human Behaviorcited by 3,568
  2. 2023MemGPT: Towards LLMs as Operating Systemscited by 1,389
  3. 2024OSWorld: Benchmarking Multimodal Agents for Open-Ended Tasks in Real Computer Environmentscited by 1,195
  4. 2024LongMemEval: Benchmarking Chat Assistants on Long-Term Interactive Memorycited by 707
  5. 2023Cognitive Architectures for Language Agentscited by 429
  6. 2024LLMLingua-2: Data Distillation for Efficient and Faithful Task-Agnostic Prompt Compressioncited by 348
  7. 2025MemOS: A Memory OS for AI Systemcited by 138
  8. 2024RAGCache: Efficient Knowledge Caching for Retrieval-Augmented Generationcited by 98
  9. 2024AIOS: LLM Agent Operating Systemcited by 97
  10. 2025SagaLLM: Context Management, Validation, and Transaction Guarantees for Multi-Agent LLM Planningcited by 60

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 4.

4 papers.

  1. 2026HarnessRisk: A Lifecycle-Oriented Benchmark for Agent Harness Safetycited by 2
  2. 2026Towards an Agent Operating System - Lessons from Classical and Cloud OScited by 1
  3. 2026The Working Set of a Coding Agent: Coherence Debt in Repository-Scale Taskscited by 1
  4. 2026AMU:Admission and Memory Update for Personalized Conversations---Structured Memory with SLM Guided Controlcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2026MemGuard: Persisting Verifier Signals for LLM-Agent Memory Governance
  2. 2026Contextual Memory Virtualisation: DAG-Based State Management and Structurally Lossless Trimming for LLM Agents
  3. 2026PLACEMEM: Toward a Compute-Aware Memory Plane for Lifelong Agents
  4. 2026Memory as Infrastructure: Reliability Engineering for Persistent Agent Memory in Months-Long LLM-Assisted Development
  5. 2026MemArchitect: A Policy Driven Memory Governance Layer
  6. 2026Shared Selective Persistent Memory for Agentic LLM Systems
  7. 2026A Survey on Long-Term Memory Security in LLM Agents: Attacks, Defenses, and Governance Across the Memory Lifecycle
  8. 2026Stale Does Not Mean Unsafe: Guard Precision for Tool-Using LLM Agents under Infrastructure State Races